Posting here because the standard appeal channel has produced no human response after three attempts, and I have run out of escalation paths. I am writing as a longtime supporter of OpenAI, not someone who is here to attack the company.
I have been a paying ChatGPT Pro subscriber since December 10, 2024. The subscription is billed monthly through Google Play and remains active. I am a heavy user of ChatGPT and Codex for my daily development.
The account was banned with an absolute block and zero prior warning. The termination email cited a generic violation of Terms and Usage Policies, with no specific clause, prompt, or behavior identified. The ban instantly revoked access to ChatGPT, Codex, API keys, and my Pro plan. Crucially, whatever I was building inside Codex is now completely inaccessible, including conversation context I cannot easily reconstruct.
For context: I am a physician and a developer. I mention this not to demand special treatment, but to clarify that I rely heavily on OpenAI’s ecosystem for legitimate work and have absolutely no incentive to abuse it.
To address the standard questions head-on:
-
Account Sharing: None. Used exclusively by me.
-
Network Masking: I do not use a VPN, proxy, or any tool to mask my IP.
-
Automation: I do not run scripts, bots, or automated scrapers against the ChatGPT UI or Codex.
-
Payments: No chargebacks or disputed payments.
-
Content: I have never attempted to jailbreak or bypass safety systems.
-
Duplicate Accounts: I have no other OpenAI accounts.
The only explanation I have for triggering this automated, zero-warning ban is my network environment. I frequently connect from the Hospital Santa Catarina Paulista network in SĂŁo Paulo during my clinical shifts.
It turns out the hospital’s enterprise network routes all its internet traffic through a Google Cloud link. I checked the IP today. The public egress IP from this clinical network resolves directly to a Google LLC Data Center/Transit IP (177.69.36.49 / IPv6 2001:4860:7:f03::de).
OpenAI’s security system must have seen my traffic coming from a Data Center IP block and automatically assumed I was a headless proxy, a bot farm, or someone masking their network. I’m not. I was literally just a regular user sitting at a hospital computer, using the corporate network. Hundreds of hospital devices share this same public IP, which clearly looks like a proxy or datacenter node to your automated fraud/abuse detection systems. I do not control this clinical network architecture and cannot change how its outbound traffic is classified.
This automated ban destroyed my development workflow and locked me out of critical code context I can’t easily reconstruct. I have zero reasons to break the rules.
I can share my User ID and Organization ID privately with any OpenAI staff member who responds here.
If anyone from the OpenAI staff is reading this, I would deeply appreciate a human review of the logs to reverse this data center IP false positive. If anyone in the community has experienced a similar absolute block due to shared corporate/hospital networks and found a resolution path, I would value your advice.
